Detection of microorganisms using terahertz ... - Semantic Scholar
OPEN SUBJECT AREAS: METAMATERIALS TERAHERTZ OPTICS
Detection of microorganisms using terahertz metamaterials S. J. Park1, J. T. Hong1, S. J. Choi2, H. S. Kim2, W. K. Park3, S. T. Han3, J. Y. Park1, S. Lee1, D. S. Kim4 & Y. H. Ahn1
SUB-WAVELENGTH OPTICS 1
Received 26 November 2013 Accepted 28 April 2014 Published 16 May 2014
Correspondence and requests for materials should be addressed to Y.H.A. (ahny@ajou. ac.kr)
Department of Physics and Division of Energy Systems Research, Ajou University, Suwon 443-749, Korea, 2Department of Biological Science, Ajou University, Suwon 443-749, Korea, 3Advanced Medical Device Research Center, Korea Electrotechnology Research Institute, Ansan 426-170, Korea, 4Center for Subwavelength Optics and Department of Physics and Astronomy, Seoul National University, Seoul 151-747, Korea.
Microorganisms such as fungi and bacteria cause many human diseases and therefore rapid and accurate identification of these substances is essential for effective treatment and prevention of further infections. In particular, contemporary microbial detection technique is limited by the low detection speed which usually extends over a couple of days. Here we demonstrate that metamaterials operating in the terahertz frequency range shows promising potential for use in fabricating the highly sensitive and selective microbial sensors that are capable of high-speed on-site detection of microorganisms in both ambient and aqueous environments. We were able to detect extremely small amounts of the microorganisms, because their sizes are on the same scale as the micro-gaps of the terahertz metamaterials. The resonant frequency shift of the metamaterials was investigated in terms of the number density and the dielectric constants of the microorganisms, which was successfully interpreted by the change in the effective dielectric constant of a gap area.
C
ulture-based detection methods such as polymerase chain reaction (PCR) and fluorescence-based microbial detection systems have been widely used in order to detect and quantify microorganisms1–3. However, although the PCR is capable of detecting a wide range of infectious fungi and bacteria, this method is timeconsuming and labor-intensive. Conversely, most of fluorescence-based microbial detection systems such as epifluorescence microscopy (EFM), flow cytometry (FC) and solid-phase cytometry (SPC) require fluorescent materials for efficient detection of the microorganisms4,5. Optical scattering methods such as light scattering and autofluorescence do not require the dyes; however they suffer from the lack of sensitivity. Therefore, there is an increasing need to develop novel techniques for the effective on-site detection of minute amounts of microbial substances. Recently, terahertz (THz) spectroscopy has emerged as a promising technique that enables the label-free, noncontact, and non-destructive inspection on the chemical and biological substances6–15. In particular, the recently developed, portable THz spectroscopic tools enable on-site detection and identification of these materials with high signal to noise ratio16,17. THz frequency range detection of microorganisms such as fungi, bacteria, and viruses has generated a great deal of interests because of its relevance to food and security inspection18. In many cases, however, the microorganisms are non-responsive in the THz frequency range since they are mostly transparent to THz waves. In addition the size of a typical microorganism is on the order of ,l/100, resulting in low scattering cross-section. Metamaterials consist of periodically arranged, sub-wavelength metallic elements and exhibit unique electromagnetic properties such as negative refraction19,20, sub-diffraction limited focusing21–23 and cloaking24,25. In addition, the metamaterials have gap structures characterized by strongly localized and enhanced fields, enabling sensitive detection of extremely small amounts of chemical and biological substances26–32. In particular, metamaterials operating in the THz frequency range have a micro-sized gap33–39 and can therefore serve as an ideal platform for the sensitive detection of fungi and bacteria, because the size of these microorganisms is compatible with the gap size. Moreover, THz metamaterials are extremely sensitive to the substances near the surface, which is favorable for sensing in aqueous environment since it allows us to use a thin water layer without suffering from the significant loss in the THz wave transmission. In this work, we performed THz time-domain spectroscopy on metamaterial sensors for the high speed detection of viable and live microorganisms such as molds, yeast cells and bacteria. We measured a frequency SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
1
www.nature.com/scientificreports shift in the inductive-capacitive resonance following the deposition of microorganisms with very low surface density. In particular, we functionalize the sensors with the antibody specific to bacteria in aqueous environment, enabling the selective detection. The resonance frequency shift in THz metamaterials was studied as a function of the dielectric constants and the density of fungi which is in accordance of the dielectric constant measurement of individual fungi. Our experimental findings are in a good agreement with the results of simulations. Schematic presentation of the experiments is shown in Fig. 1a. We measured the change in spectra of the THz radiation transmitted through the metamaterials following the deposition of microorganisms. The metamaterials consist of metallic arrays of a square ring with a micro-gap at the center. Here, the resonance of the transmission is determined both by the capacitance (C) of the gap structure and the inductance (L) associated with the shape of the square ring. In other words, pffiffiffiffiffiffithe resonance of the transmission dip is expressed by fres ~1=(2p LC ). The dielectric microorganisms placed in the gap area will cause a change in the effective dielectric constants of the capacitor, resulting in the shift of the resonant frequency in metamaterials. A representative scanning electron microscopy (SEM) image is shown in Fig. 1b, with the Penicillium chrysogenum (penicillia) deposited on the metamaterials. The fungi samples were grown by a streaking on medium method followed by the incubation at 37uC for 2 days. Metamaterial patterns were prepared by using a conventional photo-lithography method on an undoped Si substrate, followed by the metal deposition in order to define the arrays of split ring resonator patterns with a linewidth of 4 mm and gap sizes of 2–3 mm. THz transmission spectra were obtained by using the THz time-domain spectroscopic techniques with an acquisition time of 5 sec for each spectrum40,41.
Figure 1 | Sensing microorganisms using THz metamaterials. (a) A schematic presentation of THz metamaterials sensing of microorganisms. (b) A color-enhanced SEM image of metamaterials coated by penicillia. (inset) Magnified image of the fungi located in the micro-gap. SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
We first describe the THz transmission experiments for the lowdensity microorganisms that were deposited on a plain silicon substrate. Molds and Saccharomyces cerevisiae (yeasts) were deposited on the substrate by rubbing a fungi-coated swab on it. Fig. 2a shows the THz transmission amplitude obtained when the penicillia were deposited on the plain substrate at a density of 0.032/mm2, as shown by the microscope image in the inset of Fig. 2a. No noticeable change in transmission amplitude was found following the deposition of the molds compared to the bare substrate case. Many of the microorganisms such as fungi and bacteria do not have spectral fingerprints in the THz frequency range and thus are basically transparent to the THz waves. Besides, the scattering cross-section of those microbial materials to THz waves is very low, because the typical size of these microorganisms (,mm) is significantly smaller than the wavelength of THz waves. A similar behavior with respect to THz transmission was observed for other microorganisms, such as yeasts and Escherichia coli BL21(DE3) (E. coli), plated at comparable or higher densities (See Supplementary Figure S1). A representative result on the highly sensitive detection of lowdensity molds by using the THz metamaterials is demonstrated in the following. Fig. 2b shows two microscopic images of our metamaterial pattern (so-called ‘‘electrical split ring resonator’’), which is an array of a square ring with a micro-gap at the center33,34. The transmission dip yields 0.837 THz with a full-width at half maximum (FWHM) of 0.15 THz, which corresponds to the inductive-capacitive resonance for the device shown in the image. The left side image in Fig. 2b shows metamaterials with fungi (penicillia) that were deposited at a density
Figure 2 | Metamaterial sensing of penicillia. (a) THz transmission amplitudes, for transmission through a plain Si substrate, with (red solid line) and without (black dashed line) deposition of penicillia. (inset) A microscopic image of penicillia deposited on the Si substrate. (b) Two microscope images of metamaterials patterns on Si substrate, with the deposition of penicillia a (left) and after fungicide treatments (right). (c) THz transmission amplitudes measured before (gray solid line) and after (red solid line) the deposition. Transmission after the fungicide treatment is shown as blue dashed line. 2
www.nature.com/scientificreports
Figure 3 | Metamaterial sensing of E. coli in aqueous environment. (a) A schematic of selective bacteria (E. coli.) detection in solution environment. The Si substrate is coated with antibodies specific to E. coli. (b) Visible transmittance spectra of the quartz substrate coated by E. coli. (density of 0.078 mm2) both in aqueous (red line) and in ambient (black line) conditions. (c) THz transmission before (blue line) and after (red line) the deposition of E. coli. on the functionalized metamaterials in aqueous environments. (inset) A corresponding dark-field microscopic image obtained after the deposition of E. coli. (d) THz transmission before (blue line) and after (red line) the deposition of E. coli. on the sensors without the surface functionalization. (inset) A corresponding dark-field microscopic image after the deposition process.
of 0.090/mm2, and individual molds that are separated from each other can be clearly seen in this figure. As shown in the right side image, the fungi can be removed completely by using a commercial fungicide (sodium hypochlorite solution). Results for the corresponding THz transmission are shown in Fig. 2c. Unlike the results on the transmission through the plain substrate (Fig. 2a), the resonant frequency exhibited a shift towards the red (red solid line) relative to the resonant frequency obtained for clear metamaterial patterns (gray solid line). The amount of the shift was found to be 9 GHz (6% of the FWHM) for the density of penicillia shown in Fig. 2b. Once the molds were cleaned by using the fungicide, the initial resonant frequency was retrieved (blue dashed line), which verifies the reusability of our metamaterial sensors. The shift in the resonance frequency can be explained simply by the change of the dielectric environment in the gap area that modifies the capacitance according to C 5 eA/d, where A is the area, d is the distance, and e is the dielectric constant of the capacitor. It is clear that the increase in the dielectric constant due to the presence of the microorganisms results in the red-shift of the resonant frequency. Similar behaviors were also observed for yeasts and different kinds of molds, such as Neurospora sitophila (neurospora) and Aspergillus niger (niger) (See Supplementary Figure S2). We also performed experiments by using different types of the metamaterial patterns with a gap structure such as single/double split ring resonators and found similar red-shifts following the deposition of fungi; however the electrical split-ring resonator delivers better sensitivity in general (See Supplementary Figure S3). Our technique can be extended to the highly sensitive detection of bacteria in aqueous environments, which is not accessible by the conventional optical techniques. More importantly, THz metamaterials functionalized with receptors enable the specific sensing of bacteria. A representative result, demonstrating the detection of bacteria (E. coli) in a solution environment, is shown in Fig. 3. The E. coli were deposited from the solution (100 mg/ml) on the Si substrate SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
with metamaterial patterns functionalized with E. coli anti-body42, followed by a rinsing with purified water (Fig. 3a). The patterns were enclosed by a cover glass, and the water layer thickness was fixed at 55 mm by using a spacer. This thickness is large enough to cover the small detection volume that is strongly localized at the surface, yet allowing a considerable amount of THz wave transmission. We note that, the E. coli are not as clearly resolved in the solution as they are in the air. Fig. 3b shows the optical transmission spectra in the visible range, obtained for transmission through the quartz substrate coated by the E. coli (density of 0.078/mm2), for both aqueous (red line) and ambient (black line) environments. Transmission through the bare quartz substrate was used as a reference in both cases. The change in the optical transmission reaches only ,1% in water, which is 10-fold smaller as compared to the air. This strongly suggests that the microbial detection techniques based on the scattering of visible light can be limited for some of the microbial systems in the solution phase. In contrast, using the THz metamaterials sensors, we can successfully detect the low-density E. coli. Shown in Fig. 3c are the results for the THz transmission obtained with (red line) and without (blue line) the E. coli on the functionalized substrate. The inset of Fig. 3c shows a dark-field microscope image of metamaterials in water with the E. coli (density of 0.019/mm2) adsorbed on the substrate. A clear blue-shift of 23 GHz (yielding ,15% of a FWHM) is observed for the metamaterials with E. coli compared to the case of the clean substrate (blue line). The reason that the deposition of E. coli in aqueous solution causes a blue shift is likely due to the relatively low dielectric constant of E. coli (,1.6 at 1 THz) as compared to that of water (,4.2 at 1 THz)43 (Also see Supplementary Figure S4). Without the surface functionalization, however, we could not observe a noticeable change in the peak position (Fig. 3b). Therefore it is possible to fabricate the biosensors that are specific to the bacteria of interests. Quite obviously, our scheme can be extended to the selective detection of variety of microorganisms such as fungi44 and viruses 3
www.nature.com/scientificreports
Figure 4 | Resonant frequency shift as a function of number density and dielectric constants of fungi.(a) A series of six SEM images of the metamaterials with the number of penicillia in the gap areas ranging from 0 to 5. (b) THz transmission amplitudes with the penicillia number density of Navg5 0.59 and 3.4, respectively from top to bottom. (c) Same as (b) but with the yeast number density of Navg5 0.85 and 3.6, respectively from top to bottom. (d) A plot of resonant frequency shifts as a function of Navg, for depositions of penicillia (blue boxes) and yeasts (red circles). Solid lines are linear fits to the data.
since the specific binders and their surface functionalization have been widely investigated. It is evident that the shift in the resonance frequency will strongly depend on the number of fungi and bacteria located in the gap area and on the dielectric constants of the specimens. In Fig. 4, the shift of the resonant frequency is shown as a function of the number density, for different kinds of fungi. The average number of fungi in the gap area was determined by counting them for the entire set of elements (10 3 10) positioned in the spot area of the THz fields (,1 mm2). For example, Fig. 4a shows a series of SEM images of metamaterials with the number of penicillia in the gap area (,3 3 10 mm2) ranging from 0 to 5. Fig. 4b shows the representative data on the resonant frequency shifts, for two different number densities of Nav 5 0.59 (corresponding to surface density of 0.020/mm2) and 3.4 (0.113/mm2), respectively from top to bottom for the penicillia. Similar experimental data are shown in Fig. 4c for the yeast cells with Nav 5 0.85 (0.028/mm2) and 3.6 (0.120/mm2), respectively. We emphasize that we were able to detect extremely small amounts of fungi, in other words, even in the case when the average number in the gap area was less than unity. A plot of the resonant frequency shift as a function of Nav is shown in Fig. 4d for the penicillia (blue boxes) and yeasts (red SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
circles), respectively. The magnitude of the resonant frequency shift is gradually increasing with Nav for both fungi types. More importantly, it is very clear that the frequency shift is ,3-fold higher in the case of the yeasts. This is a strong indication that the dielectric constant of the yeast is higher than that of the penicillium. As mentioned earlier, the resonant frequency shift as a function of Nav and the dielectric constant of the individual fungi (ef) can be analyzed in terms of the changes in the effective dielectric constant of the capacitor. The modified effective dielectric constant e due to the presence of the fungi in the ambient condition can be expressed as e 5 eeff 1 aN(ef21), where eeff is the effective dielectric constant without the deposition of fungi, N is the number of microorganisms, and a is the coefficient which is associated with the volume fraction of the fungi. Here, eeff in the capacitance (and hence the resonant frequency) is influenced both by the dielectric constants of the substrate and the air41, in which we obtained eeff 5 6.4 from simulation results {1 (See Supplementary Figure S5). From the relation f ~f0 e=eeff 2 , where f (f0) is the resonant frequency with (without) the deposition of the fungi, the frequency shift Df 5 f 2 f0 leads to 1 Df =f0
Detection of microorganisms using terahertz metamaterials S. J. Park1, J. T. Hong1, S. J. Choi2, H. S. Kim2, W. K. Park3, S. T. Han3, J. Y. Park1, S. Lee1, D. S. Kim4 & Y. H. Ahn1
SUB-WAVELENGTH OPTICS 1
Received 26 November 2013 Accepted 28 April 2014 Published 16 May 2014
Correspondence and requests for materials should be addressed to Y.H.A. (ahny@ajou. ac.kr)
Department of Physics and Division of Energy Systems Research, Ajou University, Suwon 443-749, Korea, 2Department of Biological Science, Ajou University, Suwon 443-749, Korea, 3Advanced Medical Device Research Center, Korea Electrotechnology Research Institute, Ansan 426-170, Korea, 4Center for Subwavelength Optics and Department of Physics and Astronomy, Seoul National University, Seoul 151-747, Korea.
Microorganisms such as fungi and bacteria cause many human diseases and therefore rapid and accurate identification of these substances is essential for effective treatment and prevention of further infections. In particular, contemporary microbial detection technique is limited by the low detection speed which usually extends over a couple of days. Here we demonstrate that metamaterials operating in the terahertz frequency range shows promising potential for use in fabricating the highly sensitive and selective microbial sensors that are capable of high-speed on-site detection of microorganisms in both ambient and aqueous environments. We were able to detect extremely small amounts of the microorganisms, because their sizes are on the same scale as the micro-gaps of the terahertz metamaterials. The resonant frequency shift of the metamaterials was investigated in terms of the number density and the dielectric constants of the microorganisms, which was successfully interpreted by the change in the effective dielectric constant of a gap area.
C
ulture-based detection methods such as polymerase chain reaction (PCR) and fluorescence-based microbial detection systems have been widely used in order to detect and quantify microorganisms1–3. However, although the PCR is capable of detecting a wide range of infectious fungi and bacteria, this method is timeconsuming and labor-intensive. Conversely, most of fluorescence-based microbial detection systems such as epifluorescence microscopy (EFM), flow cytometry (FC) and solid-phase cytometry (SPC) require fluorescent materials for efficient detection of the microorganisms4,5. Optical scattering methods such as light scattering and autofluorescence do not require the dyes; however they suffer from the lack of sensitivity. Therefore, there is an increasing need to develop novel techniques for the effective on-site detection of minute amounts of microbial substances. Recently, terahertz (THz) spectroscopy has emerged as a promising technique that enables the label-free, noncontact, and non-destructive inspection on the chemical and biological substances6–15. In particular, the recently developed, portable THz spectroscopic tools enable on-site detection and identification of these materials with high signal to noise ratio16,17. THz frequency range detection of microorganisms such as fungi, bacteria, and viruses has generated a great deal of interests because of its relevance to food and security inspection18. In many cases, however, the microorganisms are non-responsive in the THz frequency range since they are mostly transparent to THz waves. In addition the size of a typical microorganism is on the order of ,l/100, resulting in low scattering cross-section. Metamaterials consist of periodically arranged, sub-wavelength metallic elements and exhibit unique electromagnetic properties such as negative refraction19,20, sub-diffraction limited focusing21–23 and cloaking24,25. In addition, the metamaterials have gap structures characterized by strongly localized and enhanced fields, enabling sensitive detection of extremely small amounts of chemical and biological substances26–32. In particular, metamaterials operating in the THz frequency range have a micro-sized gap33–39 and can therefore serve as an ideal platform for the sensitive detection of fungi and bacteria, because the size of these microorganisms is compatible with the gap size. Moreover, THz metamaterials are extremely sensitive to the substances near the surface, which is favorable for sensing in aqueous environment since it allows us to use a thin water layer without suffering from the significant loss in the THz wave transmission. In this work, we performed THz time-domain spectroscopy on metamaterial sensors for the high speed detection of viable and live microorganisms such as molds, yeast cells and bacteria. We measured a frequency SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
1
www.nature.com/scientificreports shift in the inductive-capacitive resonance following the deposition of microorganisms with very low surface density. In particular, we functionalize the sensors with the antibody specific to bacteria in aqueous environment, enabling the selective detection. The resonance frequency shift in THz metamaterials was studied as a function of the dielectric constants and the density of fungi which is in accordance of the dielectric constant measurement of individual fungi. Our experimental findings are in a good agreement with the results of simulations. Schematic presentation of the experiments is shown in Fig. 1a. We measured the change in spectra of the THz radiation transmitted through the metamaterials following the deposition of microorganisms. The metamaterials consist of metallic arrays of a square ring with a micro-gap at the center. Here, the resonance of the transmission is determined both by the capacitance (C) of the gap structure and the inductance (L) associated with the shape of the square ring. In other words, pffiffiffiffiffiffithe resonance of the transmission dip is expressed by fres ~1=(2p LC ). The dielectric microorganisms placed in the gap area will cause a change in the effective dielectric constants of the capacitor, resulting in the shift of the resonant frequency in metamaterials. A representative scanning electron microscopy (SEM) image is shown in Fig. 1b, with the Penicillium chrysogenum (penicillia) deposited on the metamaterials. The fungi samples were grown by a streaking on medium method followed by the incubation at 37uC for 2 days. Metamaterial patterns were prepared by using a conventional photo-lithography method on an undoped Si substrate, followed by the metal deposition in order to define the arrays of split ring resonator patterns with a linewidth of 4 mm and gap sizes of 2–3 mm. THz transmission spectra were obtained by using the THz time-domain spectroscopic techniques with an acquisition time of 5 sec for each spectrum40,41.
Figure 1 | Sensing microorganisms using THz metamaterials. (a) A schematic presentation of THz metamaterials sensing of microorganisms. (b) A color-enhanced SEM image of metamaterials coated by penicillia. (inset) Magnified image of the fungi located in the micro-gap. SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
We first describe the THz transmission experiments for the lowdensity microorganisms that were deposited on a plain silicon substrate. Molds and Saccharomyces cerevisiae (yeasts) were deposited on the substrate by rubbing a fungi-coated swab on it. Fig. 2a shows the THz transmission amplitude obtained when the penicillia were deposited on the plain substrate at a density of 0.032/mm2, as shown by the microscope image in the inset of Fig. 2a. No noticeable change in transmission amplitude was found following the deposition of the molds compared to the bare substrate case. Many of the microorganisms such as fungi and bacteria do not have spectral fingerprints in the THz frequency range and thus are basically transparent to the THz waves. Besides, the scattering cross-section of those microbial materials to THz waves is very low, because the typical size of these microorganisms (,mm) is significantly smaller than the wavelength of THz waves. A similar behavior with respect to THz transmission was observed for other microorganisms, such as yeasts and Escherichia coli BL21(DE3) (E. coli), plated at comparable or higher densities (See Supplementary Figure S1). A representative result on the highly sensitive detection of lowdensity molds by using the THz metamaterials is demonstrated in the following. Fig. 2b shows two microscopic images of our metamaterial pattern (so-called ‘‘electrical split ring resonator’’), which is an array of a square ring with a micro-gap at the center33,34. The transmission dip yields 0.837 THz with a full-width at half maximum (FWHM) of 0.15 THz, which corresponds to the inductive-capacitive resonance for the device shown in the image. The left side image in Fig. 2b shows metamaterials with fungi (penicillia) that were deposited at a density
Figure 2 | Metamaterial sensing of penicillia. (a) THz transmission amplitudes, for transmission through a plain Si substrate, with (red solid line) and without (black dashed line) deposition of penicillia. (inset) A microscopic image of penicillia deposited on the Si substrate. (b) Two microscope images of metamaterials patterns on Si substrate, with the deposition of penicillia a (left) and after fungicide treatments (right). (c) THz transmission amplitudes measured before (gray solid line) and after (red solid line) the deposition. Transmission after the fungicide treatment is shown as blue dashed line. 2
www.nature.com/scientificreports
Figure 3 | Metamaterial sensing of E. coli in aqueous environment. (a) A schematic of selective bacteria (E. coli.) detection in solution environment. The Si substrate is coated with antibodies specific to E. coli. (b) Visible transmittance spectra of the quartz substrate coated by E. coli. (density of 0.078 mm2) both in aqueous (red line) and in ambient (black line) conditions. (c) THz transmission before (blue line) and after (red line) the deposition of E. coli. on the functionalized metamaterials in aqueous environments. (inset) A corresponding dark-field microscopic image obtained after the deposition of E. coli. (d) THz transmission before (blue line) and after (red line) the deposition of E. coli. on the sensors without the surface functionalization. (inset) A corresponding dark-field microscopic image after the deposition process.
of 0.090/mm2, and individual molds that are separated from each other can be clearly seen in this figure. As shown in the right side image, the fungi can be removed completely by using a commercial fungicide (sodium hypochlorite solution). Results for the corresponding THz transmission are shown in Fig. 2c. Unlike the results on the transmission through the plain substrate (Fig. 2a), the resonant frequency exhibited a shift towards the red (red solid line) relative to the resonant frequency obtained for clear metamaterial patterns (gray solid line). The amount of the shift was found to be 9 GHz (6% of the FWHM) for the density of penicillia shown in Fig. 2b. Once the molds were cleaned by using the fungicide, the initial resonant frequency was retrieved (blue dashed line), which verifies the reusability of our metamaterial sensors. The shift in the resonance frequency can be explained simply by the change of the dielectric environment in the gap area that modifies the capacitance according to C 5 eA/d, where A is the area, d is the distance, and e is the dielectric constant of the capacitor. It is clear that the increase in the dielectric constant due to the presence of the microorganisms results in the red-shift of the resonant frequency. Similar behaviors were also observed for yeasts and different kinds of molds, such as Neurospora sitophila (neurospora) and Aspergillus niger (niger) (See Supplementary Figure S2). We also performed experiments by using different types of the metamaterial patterns with a gap structure such as single/double split ring resonators and found similar red-shifts following the deposition of fungi; however the electrical split-ring resonator delivers better sensitivity in general (See Supplementary Figure S3). Our technique can be extended to the highly sensitive detection of bacteria in aqueous environments, which is not accessible by the conventional optical techniques. More importantly, THz metamaterials functionalized with receptors enable the specific sensing of bacteria. A representative result, demonstrating the detection of bacteria (E. coli) in a solution environment, is shown in Fig. 3. The E. coli were deposited from the solution (100 mg/ml) on the Si substrate SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
with metamaterial patterns functionalized with E. coli anti-body42, followed by a rinsing with purified water (Fig. 3a). The patterns were enclosed by a cover glass, and the water layer thickness was fixed at 55 mm by using a spacer. This thickness is large enough to cover the small detection volume that is strongly localized at the surface, yet allowing a considerable amount of THz wave transmission. We note that, the E. coli are not as clearly resolved in the solution as they are in the air. Fig. 3b shows the optical transmission spectra in the visible range, obtained for transmission through the quartz substrate coated by the E. coli (density of 0.078/mm2), for both aqueous (red line) and ambient (black line) environments. Transmission through the bare quartz substrate was used as a reference in both cases. The change in the optical transmission reaches only ,1% in water, which is 10-fold smaller as compared to the air. This strongly suggests that the microbial detection techniques based on the scattering of visible light can be limited for some of the microbial systems in the solution phase. In contrast, using the THz metamaterials sensors, we can successfully detect the low-density E. coli. Shown in Fig. 3c are the results for the THz transmission obtained with (red line) and without (blue line) the E. coli on the functionalized substrate. The inset of Fig. 3c shows a dark-field microscope image of metamaterials in water with the E. coli (density of 0.019/mm2) adsorbed on the substrate. A clear blue-shift of 23 GHz (yielding ,15% of a FWHM) is observed for the metamaterials with E. coli compared to the case of the clean substrate (blue line). The reason that the deposition of E. coli in aqueous solution causes a blue shift is likely due to the relatively low dielectric constant of E. coli (,1.6 at 1 THz) as compared to that of water (,4.2 at 1 THz)43 (Also see Supplementary Figure S4). Without the surface functionalization, however, we could not observe a noticeable change in the peak position (Fig. 3b). Therefore it is possible to fabricate the biosensors that are specific to the bacteria of interests. Quite obviously, our scheme can be extended to the selective detection of variety of microorganisms such as fungi44 and viruses 3
www.nature.com/scientificreports
Figure 4 | Resonant frequency shift as a function of number density and dielectric constants of fungi.(a) A series of six SEM images of the metamaterials with the number of penicillia in the gap areas ranging from 0 to 5. (b) THz transmission amplitudes with the penicillia number density of Navg5 0.59 and 3.4, respectively from top to bottom. (c) Same as (b) but with the yeast number density of Navg5 0.85 and 3.6, respectively from top to bottom. (d) A plot of resonant frequency shifts as a function of Navg, for depositions of penicillia (blue boxes) and yeasts (red circles). Solid lines are linear fits to the data.
since the specific binders and their surface functionalization have been widely investigated. It is evident that the shift in the resonance frequency will strongly depend on the number of fungi and bacteria located in the gap area and on the dielectric constants of the specimens. In Fig. 4, the shift of the resonant frequency is shown as a function of the number density, for different kinds of fungi. The average number of fungi in the gap area was determined by counting them for the entire set of elements (10 3 10) positioned in the spot area of the THz fields (,1 mm2). For example, Fig. 4a shows a series of SEM images of metamaterials with the number of penicillia in the gap area (,3 3 10 mm2) ranging from 0 to 5. Fig. 4b shows the representative data on the resonant frequency shifts, for two different number densities of Nav 5 0.59 (corresponding to surface density of 0.020/mm2) and 3.4 (0.113/mm2), respectively from top to bottom for the penicillia. Similar experimental data are shown in Fig. 4c for the yeast cells with Nav 5 0.85 (0.028/mm2) and 3.6 (0.120/mm2), respectively. We emphasize that we were able to detect extremely small amounts of fungi, in other words, even in the case when the average number in the gap area was less than unity. A plot of the resonant frequency shift as a function of Nav is shown in Fig. 4d for the penicillia (blue boxes) and yeasts (red SCIENTIFIC REPORTS | 4 : 4988 | DOI: 10.1038/srep04988
circles), respectively. The magnitude of the resonant frequency shift is gradually increasing with Nav for both fungi types. More importantly, it is very clear that the frequency shift is ,3-fold higher in the case of the yeasts. This is a strong indication that the dielectric constant of the yeast is higher than that of the penicillium. As mentioned earlier, the resonant frequency shift as a function of Nav and the dielectric constant of the individual fungi (ef) can be analyzed in terms of the changes in the effective dielectric constant of the capacitor. The modified effective dielectric constant e due to the presence of the fungi in the ambient condition can be expressed as e 5 eeff 1 aN(ef21), where eeff is the effective dielectric constant without the deposition of fungi, N is the number of microorganisms, and a is the coefficient which is associated with the volume fraction of the fungi. Here, eeff in the capacitance (and hence the resonant frequency) is influenced both by the dielectric constants of the substrate and the air41, in which we obtained eeff 5 6.4 from simulation results {1 (See Supplementary Figure S5). From the relation f ~f0 e=eeff 2 , where f (f0) is the resonant frequency with (without) the deposition of the fungi, the frequency shift Df 5 f 2 f0 leads to 1 Df =f0
